TREASURY INSPECTOR GENERAL FOR TAX ADMINISTRATION

Improvements Are Needed to

Better Ensure That Refunds Claimed
on Potentially Fraudulent Tax Returns
Are Not Erroneously Released

November 12, 2015

Reference Number: 2016-40-006

This report has cleared the Treasury Inspector General for Tax Administration disclosure review process
and information determined to be restricted from public release has been redacted from this document.

Phone Number / 202-622-6500

E-mail Address / TIGTACommunications@tigta.treas.gov
Website
/ http://www.treasury.gov/tigta

HIGHLIGHTS

IMPROVEMENTS ARE NEEDED TO

BETTER ENSURE THAT REFUNDS
CLAIMED ON POTENTIALLY
FRAUDULENT TAX RETURNS ARE
NOT ERRONEOUSLY RELEASED

Highlights
Final Report issued on
November 12, 2015
Highlights of Reference Number: 2016-40-006
to the Internal Revenue Service Commissioner
for the Wage and Investment Division.

IMPACT ON TAXPAYERS
The IRSs Return Integrity and Compliance
Services organization is responsible for
identifying, evaluating, and preventing the
issuance of improper refunds. This mission
includes the protection of revenue by identifying
potentially fraudulent tax returns and verifying
the accuracy of reported income and withholding
information. The IRS reported that the Integrity
and Verification Operations function prevented
more than $15 billion in refunds for over 2 million
tax returns for confirmed identity theft or fraud
during Calendar Year 2014.

WHY TIGTA DID THE AUDIT

This audit was initiated because an IRS
employee reported to the TIGTA Office of
Investigations that the IRS was not working
some taxpayer cases in which refunds were
held. The Office of Investigations identified that
these tax accounts were not timely addressed to
ensure that the refunds are not erroneously
released. This review assessed IRS processes
to ensure that tax refunds are not erroneously
released.

WHAT TIGTA FOUND

TIGTA identified that because of a programming
error, over $27 million of refunds were
erroneously issued for 13,043 Tax Year 2013
tax returns. The programming error is overriding
the IRSs two-week processing delay on some
refund tax returns that are identified by the IRS
as potentially fraudulent. These are tax returns
that the IRS Examination function also identified

as claiming a questionable tax credit. The

portion of the refund that is not reviewed by the
Examination function is erroneously issued
before the IRS can complete its verification of
income and withholding.
TIGTA also identified that ineffective monitoring
of potentially fraudulent tax returns is resulting in
the erroneous release of refunds before the
required verification. TIGTA identified 3,910 Tax
Year 2013 tax returns selected for verification
with no indication that tax examiners verified the
returns. The IRS issued refunds totaling over
$19 million for these tax returns. The IRS did
not ensure that tax examiners timely completed
their verification work. Name mismatches in IRS
systems prevented refund holds from posting to
tax accounts. Refund holds were either not set
correctly or not functioning as intended.

WHAT TIGTA RECOMMENDED

TIGTA recommended that the IRS 1) correct the
programming that is erroneously overriding the
two-week return processing delay needed to
ensure that refunds are not issued prior to
screening and verification of the tax returns,
2) develop a periodic reconciliation process to
ensure that refunds associated with identified
potentially fraudulent tax returns are not
erroneously released, 3) develop a process to
ensure that tax examiners verify a potentially
erroneous refund claimed by a taxpayer within
the refund hold period or place an unexpiring
refund hold on the taxpayers account until
verification can be completed as required, and
4) identify why refund holds placed on some
accounts were not delaying processing of the
tax returns and address the causes.
The IRS agreed with all recommendations. The
IRS stated that it took action to address the
programming conflict that was overriding the
two-week return processing delay, plans to
periodically reconcile account inventory
information to identify refunds at jeopardy for
erroneous release, initiated actions to change
the refund hold from a temporary freeze to an
unexpiring freeze, and agreed to periodically
reconcile records between computer systems to
address anomalies and prevent the issuance of
erroneous refunds.

DEPARTMENT OF THE TREASURY

WASHINGTON, D.C. 20220

TREASURY INSPECTOR GENERAL

FOR TAX ADMINISTRATION

November 12, 2015

MEMORANDUM FOR COMMISSIONER, WAGE AND INVESTMENT DIVISION

FROM:

Michael E. McKenney
Deputy Inspector General for Audit

SUBJECT:

Final Audit Report Improvements Are Needed to Better Ensure That

Refunds Claimed on Potentially Fraudulent Tax Returns Are Not
Erroneously Released (Audit # 201540026)

This report presents the results of our review to determine if processes in the Internal Revenue
Service (IRS) Integrity and Verification Operations function ensure that tax refunds are not
erroneously released. This audit was conducted because an IRS employee reported to the
Treasury Inspector General for Tax Administration Office of Investigations that the IRS was
not working some taxpayer cases in which refunds were held. The Office of Investigations
identified that these tax accounts were not timely addressed to ensure that the refunds are not
erroneously released. This audit addresses the major management challenge of Fraudulent
Claims and Improper Payments.
Managements complete response to the draft report is included in Appendix V.
Copies of this report are also being sent to the IRS managers affected by the report
recommendations. Please contact me at (202) 622-6510 or Russell P. Martin, Assistant Inspector
General for Audit (Returns Processing and Account Services), at (978) 809-0296 if you have
questions.

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Table of Contents
Background .......................................................................................................... Page 1
Results of Review ............................................................................................... Page 5
Because of a Programming Error, Some Erroneous
Refunds Are Issued After Tax Returns Are Identified
As Potentially Fraudulent ............................................................................. Page 5
Recommendations 1 and 2: ................................................ Page 6

Ineffective Monitoring of Potentially Fraudulent Tax

Returns Resulted in the Erroneous Issuance of Refunds
Prior to Required Verification ...................................................................... Page 7
Recommendations 3 and 4: ................................................ Page 9

Appendices
Appendix I Detailed Objective, Scope, and Methodology ........................ Page 10
Appendix II Major Contributors to This Report ........................................ Page 12
Appendix III Report Distribution List ....................................................... Page 13
Appendix IV Outcome Measures............................................................... Page 14
Appendix V Managements Response to the Draft Report ....................... Page 16

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Abbreviations
BFS

Bureau of the Fiscal Service

EFDS

Electronic Fraud Detection System

IRS

Internal Revenue Service

IVO

Integrity and Verification Operations

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Background
The Internal Revenue Services (IRS) Return Integrity and Compliance Services organization is
responsible for identifying, evaluating, and preventing the issuance of improper refunds. Within
the Return Integrity and Compliance Services organization is the Integrity and Verification
Operations (IVO) function, whose mission includes support of the IRSs prerefund fraud
detection and prevention efforts (i.e., detection during tax return processing prior to a refund
being issued). The IVO function protects revenue by identifying potentially fraudulent tax
returns and verifying the accuracy of reported income and withholding information. For
example, once a potentially fraudulent tax return is identified, the IVO function screens the tax
return to determine whether verification of reported income and withholding is warranted. If
verification is warranted, the return is sent to a tax examiner in the IVO function, where this
verification is performed.
The IVO function receives its inventory from the following sources:

Dependent Database Filters The Dependent Database addresses noncompliance

relevant to the Earned Income Tax Credit and other tax benefits related to the dependency
and residency of children, various other credits, identity theft, and frivolous filers. Tax
returns meeting certain identity theft criteria are routed to the IVO functions Taxpayer
Protection Program.

Electronic Fraud Detection System (EFDS) The EFDS is the system the IRS uses to
identify potentially fraudulent paper and electronically filed tax returns. During tax
return processing, paper and electronically filed tax returns are analyzed through various
EFDS data model formulas. The data models identify suspicious paper and electronically
filed tax returns based on specific characteristics of the tax return. An associated score is
computed for each tax return. The higher the score, the greater the likelihood that the tax
return is fraudulent. Tax returns with high scores are routed to the IVO function to be
screened and verified.

External Leads The External Leads Program is an IRS program that receives external
communications (leads) about questionable tax refunds received from sources outside the
IRS such as financial institutions, Federal and State agencies, and various other
third-party providers. Leads may involve Treasury Checks, direct deposits, Automated
Clearing House deposits, refund anticipation loans, refund anticipation checks, or
third-party checks and prepaid debit cards. The IVO functions reviews these leads.

The IVO function has a total of 41 teams in nine locations (Fresno, California; Atlanta, Georgia;
Andover, Massachusetts; Kansas City, Missouri; Brookhaven, New York; Cincinnati, Ohio,
Memphis, Tennessee; Austin, Texas; and Ogden, Utah). Tax Examiners on these teams verify
Page 1

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

questionable information documents such as Form W-2, Wage and Tax Statement; Form 4852,
Substitute for Form W-2, Wage and Tax Statement; Form 1099, Information Return; and
Form 1099-R, Distributions From Pensions, Annuities, Retirement or Profit-Sharing Plans,
Individual Retirement Arrangements, and Insurance Contracts, etc.
Figure 1 shows the number of refunds that the IVO function prevented for confirmed identity
theft or fraud during Calendar Year 2014.
Figure 1: Refunds Prevented by the IVO Function for
Confirmed Identity Theft or Fraud During Calendar Year 2014
Fraud
Category

Source
Dependent Database

Identity Theft

Dollar Amount

1,071,691

$6,577,972,000

661,430

$3,960,778,704

1,733,121

$10,538,750,704

EFDS

256,065

$4,432,284,837

External Leads

194,592

$535,364,026

Subtotal

450,657

$4,967,648,863

2,183,778

$15,506,399,567

EFDS
Subtotal

Other Fraud

Returns

Total

Source: IRS Refund Fraud and Identity Theft Global Report.

Identification and verification of potentially fraudulent tax returns

Tax returns that are identified during processing as potentially fraudulent and meet IVO
workload and tolerance selection criteria are automatically resequenced (processing and issuance
of any associated refund is delayed) for two weeks. Once a questionable tax return is identified,
the tax return is placed in IVO inventory for tax examiner screening and verification.

Screening A tax examiner reviews the tax return for income and withholding
information, including comparing the information reported on the current tax year1 return
to previous tax year returns to identify inconsistencies. For returns that tax examiners
verify as legitimate, the tax examiner refiles the return (the return is considered verified
as good and the refund is to be released). If the tax examiner concludes that the tax
return is potentially fraudulent, the tax return is sent for verification and a refund hold is
placed on the individuals tax account. The refund hold will prevent the issuance of the

A 12-month accounting period for keeping records on income and expenses used as the basis for calculating the
annual taxes due. For most individual taxpayers, the tax year is synonymous with the calendar year.

Page 2

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

associated refund for 11 weeks. This hold gives the IRS additional time to verify the
legitimacy of the tax return.

Verification A tax examiner attempts to contact employers to confirm wages and

withholding reported on the potentially fraudulent tax return. If the tax examiner is
unable to verify the income with the employer, the refund is permanently frozen to
prevent it from being issued.

A prior Treasury Inspector General for Tax Administration review identified

concerns with the issuance of potentially fraudulent tax refunds due to IVO
processes
In August 2013, we reported2 that 104 tax returns sampled with refunds totaling $613,929 were
not worked timely by the IVO function. Tax examiners confirmed that 96 tax returns had false
income and withholding, but actions to prevent the issuance of the fraudulent refunds were not
taken timely. The remaining eight tax returns were not screened within the required time period
to prevent the issuance of a refund.
The IRS explained that a limit in the number of refund hold transactions that could be manually
processed each day prevented permanent holds from being placed on accounts in time to prevent
the automatic release of the refunds. The IRS indicated that changes were made in January 2013
to stop the automatic release of refunds for tax returns with confirmed false income and
withholding. The IRS stated that system capacity for pending transactions was increased
five-fold and daily manual monitoring would be performed to ensure that maximum daily limits
are not exceeded.
We recommended that the IRS ensure that actions are taken to prevent the issuance of potentially
fraudulent refunds when tax returns are not timely screened and verified and ensure that case
actions are sufficiently documented. In addition, procedures should be revised to ensure that
when tax returns identified as potentially fraudulent are also assigned to another IRS function,
the tax refunds are held until the tax return is screened and verified. IRS officials agreed with
our recommendations and planned to implement actions to extend tax account freezes to prevent
the release of potentially fraudulent tax refunds. They also planned to reemphasize
documentation requirements of case actions and revise instructions for tax examiners to require
positive verification that an issue triggering an error code or referral has been addressed by the
other function before returns are released for further processing.
This review was performed at the Wage and Investment Divisions IVO function locations in
Atlanta, Georgia, and Ogden, Utah, during the period October 2014 through July 2015. We
conducted this performance audit in accordance with generally accepted government auditing

Treasury Inspector General for Tax Administration, Ref. No. 2013-40-083, Income and Withholding Verification
Processes Are Resulting in the Issuance of Potentially Fraudulent Tax Refunds (Aug. 2013).

Page 3

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

standards. Those standards require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings and conclusions based on our
audit objective. We believe that the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objective. Detailed information on our audit
objective, scope, and methodology is presented in Appendix I. Major contributors to the report
are in Appendix II.

Page 4

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Results of Review
Because of a Programming Error, Some Erroneous Refunds Are
Issued After Tax Returns Are Identified As Potentially Fraudulent
Our review identified that a programming error is resulting in the erroneous issuance of some
refunds before the income and withholding is screened and verified. Although the IRS reported
that the IVO function prevented the issuance of more than $15 billion in refunds for confirmed
identity theft or fraud during Calendar Year 2014, we identified an additional 13,043 Tax
Year 2013 tax returns with refunds totaling over $27 million that were erroneously issued
because of the programming error. We forecast that over five years the IRS could issue
$135 million in potentially erroneous refunds due to this programming error.3
The programming error is causing the two-week resequencing marker placed on potentially
fraudulent tax returns to be overridden if the tax return is also selected by the IRS Examination
function. This results in the issuance of the portion of the associated refund that is not being
reviewed by the Examination function. Below is a hypothetical example of the error:
A tax return that includes a claim for the Earned Income Tax Credit is submitted to the
IRS for processing. During processing, the tax return is identified as having a
questionable Earned Income Tax Credit claim and is selected by the Examination
function for a prerefund review. An Examination marker is placed on the taxpayers
account indicating that the tax return is being reviewed. The portion of the refund
associated with the Earned Income Tax Credit is held from being issued. At the same
time, the tax return is sent through the EFDS, identified as potentially fraudulent,
selected for IVO screening, and a two-week return processing hold is placed on the tax
account. An IVO function tax examiner screens the tax return, concludes that it is
potentially fraudulent, and selects it for verification. The tax examiner places an
11-week refund hold on the tax account to stop the refund from issuing. However,
because of the programming error, the placement of the Examination marker on the tax
account overrides the two-week return processing hold, which results in the release of the
portion of the refund not associated with the Earned Income Tax Credit.
The IRS could have detected the erroneous issuance of refunds that we identified if it established
a process to periodically reconcile tax returns identified as potentially fraudulent between the

See Appendix IV. The five-year forecast is based on multiplying the base year by five and assumes, among other
considerations, that economic conditions and tax laws do not change.

Page 5

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

EFDS and the Integrated Data Retrieval System.4 For example, a reconciliation process would
identify tax returns identified for resequencing for which a portion of the refund was issued but
an IVO tax examiner never notated in the EFDS the refiling of the tax return (return considered
verified as good and refund to be released). The IRS stated that it completes periodic reviews for
tax returns with fraudulent characteristics to ensure that appropriate actions are taken, but it
would be difficult to perform this type of reconciliation because the IVOs EFDS inventory
system does not interface with the Integrated Data Retrieval System, which provides detailed tax
account information along with refund hold and release information.

Recommendations
The Commissioner, Wage and Investment Division, should:
Recommendation 1: Correct the programming that is erroneously overriding the IVO
function resequencing tax account marker to ensure that refunds are not issued prior to tax
examiners screening and verifying the tax return.
Managements Response: The IRS agreed with this recommendation. The IRS
agreed there is a programming conflict causing an erroneous override of the IVO return
resequencing function. In June 2015, the IRS took action to reflect current processing
conditions and prevent a conflict of business rules. A programming update request was
input for the 2016 Filing Season which will resolve the conflict; however, the
implementation of such programming changes are subject to budgetary constraints,
limited resources, and competing priorities. Consequently, and due solely to those
constraints, the IRS cannot provide an implementation date.
Recommendation 2: Develop a process that periodically reconciles IVO function inventories
between the EFDS and the Integrated Data Retrieval System to ensure that refunds associated
with identified potentially fraudulent tax returns are not erroneously released.
Managements Response: The IRS agreed with this recommendation. The IRS will
periodically reconcile information from the EFDS to Master File data to identify refunds
at jeopardy for erroneous release.

IRS computer system capable of retrieving or updating stored information. It works in conjunction with a
taxpayers account records.

Page 6

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Ineffective Monitoring of Potentially Fraudulent Tax Returns Resulted

in the Erroneous Issuance of Refunds Prior to Required Verification
Our review identified 3,910 Tax Year 2013 tax returns the IRS selected for IVO verification
as being potentially fraudulent with no indication that tax examiners verified the income,
withholding, and other information as required. The IRS issued refunds totaling over
$19 million for these returns. We forecast that over five years the IRS could issue more than
$95 million in potentially erroneous refunds because tax returns are not being verified as
required.5
When we brought this issue to IVO managements attention, they stated that the erroneous
release of the refunds associated with these 3,910 tax returns resulted from:

The IRS not having an effective process to ensure that tax examiners verify a potentially
erroneous refund claimed by a taxpayer within the 11-week hold period or place an
unexpiring marker on the taxpayers account to hold the refund longer until verification
can be completed as required. For example, management does not monitor tax returns as
the 11-week refund hold expiration date approaches to ensure that the return is verified as
required or that an additional refund hold (the IRS refers to this additional refund hold as
an unexpiring refund hold) is placed on the tax account. When IVO tax examiners do
not take timely actions to verify the tax return information during the 11-week hold or
place an additional refund hold marker on the tax account prior to expiration of the
11-week hold, the refund will be systemically released and issued.

Name mismatches between the EFDS and the IRSs Master File. IVO officials stated that
their process to place unexpiring refund holds on some tax accounts requires analysts to
record names and other information from the EFDS to a spreadsheet and upload the
spreadsheet to the Master File. However, when the names on the spreadsheet are not an
exact match to the Master File, the refund holds will not post to the tax accounts and the
refunds will be erroneously issued. Officials stated that they implemented a process in
March 2015 to ensure that the IRS identifies these name mismatches and inputs the
refund holds. They also stated that the IRS now identifies these returns within 10 weeks
of the 11-week refund hold. The related accounts are then researched and corrections are
manually uploaded to the Master File on a weekly basis.
The name mismatches are similar to another name mismatch issue that we reported in
September 2014.6 In this prior review, we reported that some tax returns in the EFDS had

See Appendix IV. The five-year forecast is based on multiplying the base year by five and assumes, among other
considerations, that economic conditions and tax laws do not change.
6
Treasury Inspector General for Tax Administration, Ref. No. 2014-40-091, Prisoner Tax Refund Fraud: Delays
Continue in Completing Agreements to Share Information With Prisons, and Reports to Congress Are Not Timely or
Complete (Sept. 2014).

Page 7

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

a two-character name control instead of the four-character name control in the Master
File. A name control is the first four letters in an individuals last name, e.g., the name
control for Smith would be SMIT, but the two-character field would be something
unrelated to the name control that loaded into the EFDS. The IRSs review of these tax
returns identified that the two-character name control (included on preprinted address
labels provided to the taxpayer by the IRS) was being loaded into the EFDS instead of the
actual name control. This deficiency that we reported in our prior review resulted in the
IRS not assigning a prisoner indicator to some tax accounts as warranted. We
recommended that the IRS correct the computer programming errors that caused this
deficiency. The IRS did not agree with this recommendation.

Tax returns being routed directly to the IVO scheme tracking system7 during the same
time frame that they were received in the EFDS. These tax returns involve known tax
fraud schemes, and when tax returns associated with these schemes are processed, they
are identified and routed directly to the IVO scheme tracking system. However, the
direct routing requires a manual process to add an unexpiring refund hold marker to the
tax account. IVO management indicated that the unexpiring refund hold should be added
before the tax return is routed to its scheme tracking system. However, the unexpiring
refund hold was not input on the tax accounts we identified, resulting in the erroneous
issuance of the refunds. IRS officials stated that this weakness has been mitigated
because tax returns routed directly to its scheme tracking system now first have an
11-week refund hold placed on the tax account.

Resequencing markers systemically set on some tax accounts did not delay processing the
tax returns, which resulted in the release of the associated refunds. IVO officials
informed us they are investigating this problem.

Unexpiring refund hold markers were not placed on taxpayers accounts by the IVO
function. These refunds were identified by the Dependent Database filters as suspicious
because taxpayers claimed questionable credits on their tax returns. However, in
Calendar Year 2014, the tax accounts needing an unexpiring account marker were
manually recorded on a spreadsheet, which was then used to input the marker to the
Master File. This process was not always effective. Thus, some tax accounts were not
updated with the refund holds, and the refunds were issued before the IVOs Automated
Questionable Credits program8 could verify taxpayers eligibility for the credits. IVO
officials stated that this problem is now resolved because the manual process was
discontinued and the markers are now systemically input.

The Scheme Tracking and Referral System is the subsystem of the EFDS that captures statistical information and
tracks the status of refund schemes.
8
Tax examiners in the Automated Questionable Credits program review tax accounts and determine if appropriate
documentation exists for the credit(s) claimed.

Page 8

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Recommendations
The Commissioner, Wage and Investment Division, should:
Recommendation 3: Develop a process to ensure that tax examiners verify a potentially
erroneous refund claimed by a taxpayer within the 11-week hold period or place an unexpiring
refund hold marker on the taxpayers account to hold the refund longer until verification can be
completed as required.
Managements Response: The IRS agreed with this recommendation. The IRS
agreed that returns should be verified or an unexpiring refund hold input on a taxpayers
account prior to the 11-week refund hold. The IRS currently has a manual monitoring
process in place to use queries of the EFDS to identify accounts that need additional
refund hold actions. A programming update was requested for the 2016 Filing Season to
change the initial systemic refund freeze action from a temporary freeze to an unexpiring
freeze; however, the implementation of such programming changes are subject to
budgetary constraints, limited resources, and competing priorities. Consequently, and
due solely to those constraints, the IRS cannot provide an implementation date.
Recommendation 4: Ensure that actions are taken to identify and address the causes for the
instances in which resequencing markers placed on accounts were not delaying the IRSs
processing of the tax returns and thus preventing the erroneous issuance of refunds.
Managements Response: The IRS agreed with this recommendation. The IRS will
complete periodic reconciliations of the EFDS data to Master File information to ensure
that resequence markers are working appropriately and actions are taken to address
anomalies to prevent erroneous issuance of refunds.

Page 9

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Appendix I

Detailed Objective, Scope, and Methodology

Our overall objective was to assess the IVO function processes to ensure that tax refunds are not
erroneously released.
I.

Evaluated IVO processes to ensure that refunds are not erroneously released for
potentially fraudulent tax returns without being screened and verified by the IVO
function.
A. Identified the universe of Tax Year1 2013 returns routed to the IVO function for
screening and verification in Processing Year2 2014.
B. Identified the population of returns in the Step I.A universe for which a refund was
issued and confirmed by the Bureau of the Fiscal Service. We created a list of these
returns.
C. Identified Tax Year 2013 accounts for the scenario discovered in which the
Examination function indicator overrode an IVO indicator. The Examination
function indicator caused a portion (usually the Earned Income Tax Credit) of the
taxpayers refund claim to freeze. Any remaining refund claimed (prior to the IVO
function having a chance to review this portion of the refund) was immediately
released.
D. Validated the list of returns routed to the IVO function with a confirmed refund from
Step I.B by comparing data fields from 20 records on this list to the Integrated Data
Retrieval System.3
E. Validated the EFDS data extract by comparing a judgmental sample4 of 20 returns in
the extract to the EFDS data online. For each return sampled, we ensured that all
requested data fields were in the extract.
F. Removed from our list all returns with indications that a tax examiner worked the
case prior to the release of the refund.

A 12-month accounting period for keeping records on income and expenses used as the basis for calculating the
annual taxes due. For most individual taxpayers, the tax year is synonymous with the calendar year.
2
The calendar year in which a tax return or document is processed by the IRS.
3
IRS computer system capable of retrieving or updating stored information. It works in conjunction with a
taxpayers account records.
4
A judgmental sample is a nonprobability sample, the results of which cannot be used to project to the population.

Page 10

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

G. Provided the list from Step I.F of potentially erroneous refunds to IVO officials for
review and explanation.
1. Interviewed IVO officials to determine the reasons the refunds were issued
without being worked.
2. Assessed the validity of the reasons provided by IVO officials and performed
additional queries, as needed, to refine our query and create a final list of accounts
with erroneous refunds that were issued without being worked.
3. For the final list of erroneous refunds, we interviewed IVO officials to obtain
agreement and determined the reasons for potentially erroneous refunds.
H. Interviewed management to determine the reasons why returns failed to resequence
and the effect on refunds. A contract statistician assisted with reviewing outcome
measure projections.
Internal controls methodology
Internal controls relate to managements plans, methods, and procedures used to meet their
mission, goals, and objectives. Internal controls include the processes and procedures for
planning, organizing, directing, and controlling program operations. They include the systems
for measuring, reporting, and monitoring program performance. We determined that the
following internal controls were relevant to our audit objective: 1) the processing controls to
ensure that the IRS can identify and track tax returns identified for verification by the IVO
function using the EFDS and 2) the controls that prevent refunds for these tax returns from being
issued prior to IVO verification. We evaluated these controls by assessing the EFDS data
interface used by IVO tax examiners, evaluating the EFDS data supplied by the IVO function to
perform our data queries, and by comparing the EFDS data results to the Integrated Data
Retrieval System.

Page 11

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Appendix II

Major Contributors to This Report

Russell Martin, Assistant Inspector General for Audit (Returns Processing and Account Services)
Allen Gray, Director
Paula Johnson, Audit Manager
Tracy Harper, Lead Auditor
Tanya Boone, Senior Auditor
Lynn Faulkner, Senior Auditor
Laura Haws, Senior Auditor
Jerome Antoine, Auditor
James Allen, Information Technology Specialist

Page 12

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Appendix III

Report Distribution List

Commissioner C
Office of the Commissioner Attn: Chief of Staff C
Deputy Commissioner for Operations Support OS
Deputy Commissioner for Services and Enforcement SE
Deputy Commissioner, Wage and Investment Division SE:W
Director, Customer Account Services, Wage and Investment Division SE:W:CAS
Director, Strategy and Finance, Wage and Investment Division SE:W:S
Senior Operations Advisor, Wage and Investment Division SE:W:S
Director, Customer Account Services, Accounts Management, Wage and Investment Division
SE:W:CAS:AM
Director, Customer Account Services, Submission Processing, Wage and Investment Division
SE:W:CAS:SP
Chief Counsel CC
National Taxpayer Advocate TA
Director, Office of Program Evaluation and Risk Analysis RAS:O
Director, Office of Audit Coordination OS:PPAC:AC
Office of Internal Control OS:CFO:CPIC:IC
Audit Liaison: Chief, Program Evaluation and Improvement, Wage and Investment Division
SE:W:S:PEI

Page 13

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Appendix IV

Outcome Measures
This appendix presents detailed information on the measurable impact that our recommended
corrective actions will have on tax administration. These benefits will be incorporated into our
Semiannual Report to Congress.

Type and Value of Outcome Measure:

Revenue Protection Potential; $135,456,560 in erroneous refunds over five years for
65,215 tax returns for which the IRSs two-week tax return resequencing marker is
overridden by a marker that the Examination function uses to select tax returns for review of
questionable refundable credits (see page 5).

Methodology Used to Measure the Reported Benefit:

We used computer analysis to identify 13,043 Tax Year 2013 tax returns with refunds totaling
$27,091,312 that should not have been issued. These tax returns were identified as potentially
fraudulent with a resequencing marker added to the account to delay processing the tax returns
and keep the refunds from being issued. However, a programming change in January 2008
resulted in the markers being overridden and portions of the refunds being erroneously released.
We forecast that the IRS could potentially pay $135,456,560 in erroneous refunds for 65,215 tax
returns over the next five years. We computed this forecast by multiplying the Tax Year 2013
total erroneous refund amount ($27,091,312) by five years and the related Tax Year 2013 tax
returns (13,043) by five years.1
The actual amount of erroneous refunds paid is contingent upon the verification of income and
withholding information shown on each tax return.

The five-year forecast is based on multiplying the base year by five and assumes, among other considerations, that
economic conditions and tax laws do not change.

Page 14

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Type and Value of Outcome Measure:

Revenue Protection Potential; $95,897,105 in erroneous refunds over five years issued for
19,325 tax returns selected for IVO review but not screened and verified by IVO screeners
(see page 7).

Methodology Used to Measure the Reported Benefit:

We used computer analysis to identify 3,910 Tax Year 2013 tax returns selected for IVO review
and verification but with no indication that tax examiners screened and verified the returns.
First, we identified the universe of Tax Year 2013 tax returns routed to the IVO function for
screening and verification in Processing Year2 2014. We then used data in the IRSs Integrated
Data Retrieval System and EFDS to identify the erroneous refunds.
Of these erroneous refunds, 45 were due to a specific problem that IVO officials stated has been
resolved and are thus not included in this forecast. We forecast that the IRS could potentially
pay $95,897,105 in erroneous refunds for 19,325 tax returns over the next five years. We
computed this forecast by multiplying the total erroneous refund amount for Tax Year 2013
($19,179,421) and the 3,865 related tax returns by five years.3
The actual amount of erroneous refunds paid is contingent upon the verification of income and
withholding information shown on each tax return.

The calendar year in which a tax return or document is processed by the IRS.
The five-year forecast is based on multiplying the base year by five and assumes, among other considerations, that
economic conditions and tax laws do not change.
3

Page 15

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Appendix V

Managements Response to the Draft Report

Page 16

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Page 17

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Page 18

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Page 19

Improvements Are Needed to Better Ensure

That Refunds Claimed on Potentially Fraudulent
Tax Returns Are Not Erroneously Released

Page 20